U.S. patent number 10,320,153 [Application Number 14/714,248] was granted by the patent office on 2019-06-11 for systems and methods of an l-switched light emission.
The grantee listed for this patent is Rudolf Heinrich Binder, Nai-Hang Kwong, Paul Bryan Lundquist. Invention is credited to Rudolf Heinrich Binder, Nai-Hang Kwong, Paul Bryan Lundquist.
![](/patent/grant/10320153/US10320153-20190611-D00000.png)
![](/patent/grant/10320153/US10320153-20190611-D00001.png)
![](/patent/grant/10320153/US10320153-20190611-D00002.png)
![](/patent/grant/10320153/US10320153-20190611-D00003.png)
![](/patent/grant/10320153/US10320153-20190611-D00004.png)
![](/patent/grant/10320153/US10320153-20190611-D00005.png)
![](/patent/grant/10320153/US10320153-20190611-D00006.png)
![](/patent/grant/10320153/US10320153-20190611-D00007.png)
![](/patent/grant/10320153/US10320153-20190611-D00008.png)
![](/patent/grant/10320153/US10320153-20190611-D00009.png)
![](/patent/grant/10320153/US10320153-20190611-D00010.png)
View All Diagrams
United States Patent |
10,320,153 |
Binder , et al. |
June 11, 2019 |
Systems and methods of an L-switched light emission
Abstract
Provided herein are systems and methods for switching the
generation of light emissions using charge separation in a gain
medium to manipulate carrier lifetimes. For a given output pulse
energy, extended carrier lifetimes may allow carrier generation
powers to be reduced and/or carrier generation times to be
extended. L-switching of light output from a gain medium may be
combined with other switching schemes utilizing different
approaches to control lasing, such as Q-switching.
Inventors: |
Binder; Rudolf Heinrich
(Tucson, AZ), Kwong; Nai-Hang (Tucson, AZ), Lundquist;
Paul Bryan (Longmont, CO) |
Applicant: |
Name |
City |
State |
Country |
Type |
Binder; Rudolf Heinrich
Kwong; Nai-Hang
Lundquist; Paul Bryan |
Tucson
Tucson
Longmont |
AZ
AZ
CO |
US
US
US |
|
|
Family
ID: |
66767443 |
Appl.
No.: |
14/714,248 |
Filed: |
May 15, 2015 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
62025005 |
Jul 15, 2014 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H01S
5/04253 (20190801); H01S 5/3214 (20130101); H01S
5/34333 (20130101); H01S 5/04256 (20190801); H01S
5/3415 (20130101); H01S 5/0614 (20130101); H01S
5/041 (20130101); H01S 5/02484 (20130101); H01S
5/0424 (20130101); H01S 5/02469 (20130101); H01S
5/4043 (20130101); H01S 5/305 (20130101); H01S
5/3054 (20130101); H01S 5/026 (20130101); H01S
5/3416 (20130101) |
Current International
Class: |
H01S
5/06 (20060101); H01S 3/0933 (20060101); H01S
5/343 (20060101); H01S 5/042 (20060101); H01S
5/062 (20060101); H01S 3/0941 (20060101); H01S
3/102 (20060101); H01S 3/11 (20060101) |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
Yamanishi et al. ("Quantum mechanical size effect modulation light
sources--a new field effect semiconductor laser or light emitting
device", Jap. Jour. Applied Phys., vol. 22, No. 1, Jan. 1983, pp.
L22-L24). cited by examiner .
Takeoka et al. ("A 140ps optical pulse generation by field-induced
gain switching in a photo-excited quantum well laser", Jap. J.
Applied Phys., vol. 26, No. 2, Feb. 1987, pp. L117-L119. cited by
examiner .
Suemune et al., "Gain-Switching Characteristics and Fast Transient
Response of Three-Terminal Size-Effect Modulation Laser," IEEE
Journal of Quantum Electronics, vol. QE-22, No. 9, Sep. 1986. cited
by applicant.
|
Primary Examiner: Van Roy; Tod T
Attorney, Agent or Firm: Furman IP Law
Government Interests
STATEMENT OF GOVERNMENT SPONSORSHIP
This invention was made with government support under contract #
N00014-13-P-1173 awarded by the Office of Naval Research. The
government has certain rights in the invention.
Parent Case Text
RELATED APPLICATIONS
The application claims priority to U.S. Provisional Patent
Application Ser. No. 62/025,005, entitled "SYSTEMS AND METHODS FOR
AN L-SWITCHED LIGHT EMISSION," and filed Jul. 15, 2014, the
disclosure of which is hereby incorporated by reference in its
entirety for all purposes.
Claims
What is claimed is:
1. A light emitter comprising: a gain structure for providing an
optical gain via stimulated emission, the gain structure including
a confinement region for charge carriers within a direct-bandgap
structure defined by a continuum of charge carrier densities of
states along each of a first dimension and a second dimension, the
first dimension and the second dimension comprising a planar region
of the confinement region, the continuum of charge carrier
densities of states adapted for accumulating charge carriers in the
direct-bandgap structure, wherein the direct-bandgap structure is
configured for producing a stimulated pulsed-light emission
comprising a multitude of photons; an input surface of the light
emitter configured to route light to the confinement region where
the light undergoes the optical gain via stimulated emission; an
optical emission path of the light emitter that is orthogonal the
planar region of the confinement region and that directs light
created via stimulated emission in the gain structure through an
emission surface of the light emitter; and a plurality of
electrodes configured to modify an electric field in a direction
orthogonal to the planar region of the confinement region to
spatially-redistribute charge carriers within the direct-bandgap
structure into at least: A) a spatially-accumulating configuration
of the charge carriers that does not result in the stimulated
pulsed-light emission from the confinement region, and that: 1.
accumulates charge carriers into the continuum of charge carrier
densities of states in the direct-bandgap structure, while the
direct-bandgap structure is in a direct-bandgap state, in the
confinement region; and 2. distributes the accumulated charge
carriers, of both polarities, into portions of the direct-bandgap
structure in the confinement region; and B) a spatially-overlapping
configuration of the accumulated charge carriers for depletion of
the accumulated charge carriers that have been accumulated in the
spatially-accumulating configuration in the continuum of charge
carrier densities of states in the direct-bandgap structure in the
confinement region, wherein the spatially-overlapping configuration
spatially-distributes the accumulated charge carriers to produce
the stimulated pulsed-light emission comprising the multitude of
photons from the confinement region.
2. The light emitter of claim 1, wherein the electric field is due
in part to a built-in voltage gradient within the gain structure at
a zero bias voltage along the confinement dimension.
3. The light emitter of claim 1, wherein the electric field is due
in part to a non-zero bias voltage on the electrodes.
4. The light emitter of claim 1, further comprising: a pumping
source configured to accumulate the charge carriers over a period
of time into the spatially-accumulating configuration in the
confinement region.
5. The light emitter of claim 4, wherein the pumping source is an
optical pumping source of pump light, configured to populate the
charge carriers into the spatially-accumulating configuration in
the direct bandgap structure by absorption of the pump light by the
light emitter.
6. The light emitter of claim 5, wherein the optical pumping source
is a pump diode.
7. The light emitter of claim 1, further comprising: a carrier
injection source for populating the charge carriers in the
spatially-accumulating configuration in the confinement region.
8. The light emitter of claim 7, wherein the carrier injection
source is a heterojunction.
9. The light emitter of claim 1, wherein the light emitter is part
of a laser system including an optical laser cavity.
10. The light emitter of claim 1, wherein the optical emission path
traverses at least one of the plurality of electrodes for producing
the electric field orthogonal to the planar region of the
confinement region and, wherein the at least one of the plurality
of electrodes is at least partially transmissive.
11. The light emitter of claim 10, wherein the optical emission
path traverses at least two electrodes of the plurality of
electrodes that are each at least partially transmissive.
12. The light emitter of claim 10, wherein the light emitter is
configured with a reflective surface to create a reflected light
path including the optical emission path through the at least one
of the plurality of electrodes.
13. The light emitter of claim 1, wherein the emission surface of
the light emitter and the input surface of the light emitter
comprise a single surface of the light emitter.
14. The light emitter of claim 1, wherein the emission surface of
the light emitter is a different surface of the light emitter than
the input surface of the light emitter.
Description
FIELD OF THE TECHNOLOGY
Embodiments of this disclosure relate to optical gain structures
with modified carrier lifetimes created via using an arrangement of
quantum structures and a newly-introduced L-switching scheme
utilizing an electrical field across the quantum structures.
SUMMARY OF THE DESCRIPTION
Provided herein are systems and methods for switching the
generation of light emissions using charge separation in a gain
medium to manipulate carrier lifetimes. For a given output pulse
energy, extended carrier lifetimes may allow carrier generation
powers to be reduced and/or carrier generation times to be
extended. L-switching of light output from a gain medium may be
combined with other switching schemes utilizing different
approaches to control lasing, such as Q-switching.
In one aspect, the disclosure describes a light emitter comprising
an optical gain structure including a confinement region for charge
carriers. The light emitter further comprises a plurality of
electrodes configured to provide an electric field across the
region and produce an electric field within the region while the
electrodes maintain a first bias voltage. The light emitter further
includes a pumping source configured to populate charge carriers
into a barrier region of the quantum confined structure. The light
emitter further includes control circuitry for switching the
electric field via changing the first bias voltage to a second bias
voltage and thereby inducing recombination of the charge carriers
in the gain structure.
The light emitter may be further configured within an optical
cavity to produce a laser.
Other embodiments and features of the present disclosure will be
apparent from the accompanying drawings and from the detailed
description which follows.
BRIEF DESCRIPTION OF THE DRAWINGS
The following patent description and drawings are illustrative and
are not to be construed as limiting.
FIG. 1 shows an exemplary embodiment of a structure and related
band diagram of pumping/emission scheme for producing an L-Switched
light emission as provided for herein.
FIG. 2 shows a band diagram during a pumping/shelving phase of an
exemplary embodiment of an L-Switched light emission.
FIG. 3 shows a band diagram during an emission phase of an
exemplary embodiment of an L-Switched light emission.
FIG. 4 shows an exemplary embodiment of a structure for producing
an L-Switched light emission.
FIG. 5 shows further details of an exemplary embodiment of a
structure for producing an L-Switched light emission.
FIG. 6 shows an exemplary block diagram of a quantum well/p-n
junction structure for producing an L-Switched light emission.
FIG. 7A shows an exemplary band diagram of a quantum well/p-n
junction structure during a shelving phase.
FIG. 7B shows an exemplary band diagram of a quantum well/p-n
junction structure during an emission phase.
FIG. 8 shows a comparison across an exemplary quantum well/barrier
structure of band diagrams and strain profile of the structure.
FIG. 9 shows an exemplary band diagram of a quantum well/p-n
junction structure with the effect of charge distribution in the
junction's depletion region.
FIG. 10 shows exemplary wavefunctions within a quantum well/barrier
structure during a shelving phase.
FIG. 11 shows exemplary wavefunctions within a quantum well/barrier
structure during an emission phase.
FIG. 12 shows an exemplary band diagram of a quantum well/barrier
structure and non-radiative Shockley-Read-Hall (SRH) recombination
processes therein.
FIG. 13 shows an exemplary energy diagram showing Auger
recombination processes therein.
FIG. 14 shows an exemplary recombination factor for non-radiative
Shockley-Read-Hall (SRH) recombination processes within a quantum
well/barrier structure during the shelving phase.
FIG. 15 shows an exemplary recombination factor for non-radiative
SRH recombination processes within a quantum well/barrier structure
during the emission phase.
FIG. 16A shows a schematic of an electron wavefunction during
shelving.
FIG. 16B shows a schematic of the electron capture after
switching.
FIG. 17A shows a band diagram schematic of carrier leakage due to
tunneling in an exemplary structure.
FIG. 17B shows estimated calculations of tunneling time in such an
exemplary structure.
FIG. 18 shows an exemplary embodiment of band diagram showing
carrier energies as functions of momentum in the plane of the
quantum well.
FIG. 19 shows output power of a light pulse from an exemplary
L-switched laser.
FIG. 20A shows shelved electron/hole pair density as a function of
time.
FIG. 20B shows population inversion at the laser frequency as a
function of time.
FIG. 21A shows additional detail of cavity photon density and
electron/hole pair density around the switching time.
FIG. 21B shows additional detail of population inversion at the
laser frequency around the switching time.
FIG. 22 shows a graph of exemplary population inversion at the
laser frequency versus the shelved electron/hole pair density.
FIG. 23 shows a graph of pumping efficiency versus switching
time.
FIG. 24A shows a graph of switching time versus absorbed current
density.
FIG. 24B shows analytic estimates of efficiencies for the noted
efficiencies for an L-switched emission versus absorbed current
density (solid lines), as compared with simulation results
(squares).
FIG. 25A shows an exemplary embodiment of a multiple layer emitting
device with interdigitized electrodes from a top view.
FIG. 25B shows an exemplary embodiment of a multiple layer emitting
device with interdigitized electrodes from a side view.
FIG. 26A shows an exemplary embodiment of a structure with multiple
gain elements within a cavity from a side view.
FIG. 26B shows an exemplary embodiment of a structure with multiple
gain elements within a cavity from a top view.
FIG. 27A shows an exemplary embodiment of a structure with multiple
gain elements within a single integrated layer as a synchronized
linear array.
FIG. 27B shows an exemplary embodiment of a structure with multiple
gain elements arranged as displaced tiers.
FIG. 28 shows an exemplary structure for zero-dimensional and
one-dimensional quantum confined volumes.
FIG. 29 shows an exemplary gain stack with multiple quantum wells
per electrode pair.
FIG. 30 shows an exemplary configuration for using multiple
transmissive gain stacks in succession.
FIG. 31 shows an exemplary configuration for using multiple
reflective gain stacks in succession.
FIG. 32 shows an exemplary configuration for using side-pumping
diodes with an integrated switched gain element.
FIG. 33 shows an exemplary configuration for using external optical
pumping with a gain element.
DESCRIPTION
The following patent description and drawings are illustrative and
are not to be construed as limiting. Numerous specific details are
described to provide a thorough understanding. However, in certain
instances, well-known or conventional details are not described in
order to avoid obscuring the description. References to one or an
embodiment in the present disclosure are not necessarily references
to the same embodiment; and, such references mean at least one.
Reference in this specification to "one embodiment" or "an
embodiment" or the like means that a particular feature, structure,
or characteristic described in connection with the embodiment is
included in at least one embodiment of the disclosure. The
appearances of the phrase "in one embodiment" or the like in
various places in the specification are not necessarily all
referring to the same embodiment, nor are separate or alternative
embodiments mutually exclusive of other embodiments. Moreover,
various features are described that may be exhibited by some
embodiments and not by others.
Described herein is a pulsed semiconductor light emission device
that can be pumped over much longer durations than the emission
time via manipulating the energetic fields within the device. This
capability minimizes pumping power constraints for a given output
peak power by allowing for longer pumping times to be used for a
given pumping energy input.
Q-switched diode-pumped solid state laser architectures are
presently the primary means for generating high peak power pulsed
laser sources. These solid state lasers operate by allowing the
pumping phase to occur over much longer durations than the optical
emission phase. The energy stored within the solid state crystal's
upper energy states is built during a long pumping time, allowing
for high peak powers to be produced during emission. However, most
solid state lasers crystals only lase efficiently within a few
wavelengths of the energy bands defined by the solid state crystal
materials.
In contrast, semiconductor quantum well light emitters can be
created with quantum-confined structures designed to produce
emissions over very broad wavelength bands, from UV to the
infrared. For example, the properties of a quantum well determines
the lowest energy states that electrons (and holes) can have
therein and, thus, the emission wavelengths produced when the
carriers in the device recombine. Consequently, designed emission
wavelengths are achieved via adjusting properties such as quantum
well material and the quantum well thickness. Similar emission
wavelength selection is also possible in quantum rod geometric
designs.
However, semiconductor lasers are not used for high peak power
because the carrier recombination lifetimes in semiconductor gain
media are naturally very short. In other words, without
intervention via the methods and structures described herein,
emissions occur in semiconductor gain media at nearly the same rate
as pumping can occur, and therefore, there is no substantial energy
storage possible within the crystal and high peak power switched
laser emission operation is impossible.
The systems and methods described herein use semiconductors
directly to produce high peak power through nanoscale energy
storage using quantum structures and manipulated energy bands
therein. These systems and methods allow pumping (e.g., via optical
pumping, via heterojunction carrier injection) to occur more
slowly. This reduces the impact of various constraints including
heat production, cooling, pumping power needed in light emitting
systems, particularly in compact laser systems with multiple such
constraints. At the transition between pumping phase and emission
phase, as internal fields are released, carriers in the device need
only move several nanometers to recombine. Therefore, in addition
to the benefits of increased pumping time, in certain embodiments,
negligible heat is produced during the carrier migration necessary
to recombine.
The systems and methods described herein may use electrodes and
structural spacings to enable large electric field modulation to be
produced across the nanoscale and/or quantum structures described
herein. In configurations using optical pumping, cladding layers
may be used in the device to mitigate or prohibit carrier transport
out of the device.
In one embodiment, the device utilizes an externally applied
electric field to modulate electron and hole recombination rates.
During pumping, electric fields separate electrons and holes so
that their probability of recombination is very low. Electric
fields may be placed across the entire device and electric fields
on a nanoscale may be modified by carrier population differences,
e.g., across the distances of nanoscale structures in the
device.
In some embodiments, quantum-confined structures may be embedded
within heterojunctions such as p-n junctions. Stacked devices may
be constructed, including cladded quantum-confined structures
separated by interdigitized electrodes.
During the pumping phase of operation, carriers (e.g., electrons
and holes) are stored in the modified band structures of the
nanoscale structures. The electric field in the device is modified
abruptly to quickly shift the recombination rates of the carriers
and create a pulsed optical emission. The modification of the
electric field brings electrons and holes out of the barrier
regions and into one or more quantum-confined volumes where
recombination of the carriers occurs at high probability and thus
leading to a pulsed emission of light at the designed wavelength.
The recombination rates resulting in such a pulsed system can
result in a gain exceeding the cavity lasing threshold, resulting
in pulses with high peak power pulses.
Such a system of operation is termed herein an L-switching
switching scheme with reference to the recombination lifetime of
carriers within the device. This name invokes the terminology of an
actively Q-switched laser, only in an L-switching scheme the
emission/recombination lifetime of the carriers is modulated
instead of the cavity Q being modulated to create the output
pulse.
The two switching schemes, L-switching and Q-switching, may also be
combined into a hybrid LQ-switching scheme. During the electrical
modulation of the device, the emission spectrum will change, so
intra-cavity insertion of a spectral filter can also be used to
induce Q-switching behavior into the cavity at the same time.
Thereby two different mechanisms may be employed to hold off lasing
within the laser cavity. Thus such a hybrid system of operation is
termed herein as an LQ-switching switching scheme.
Devices containing nanostructures for containing and delaying
recombination of carriers may be constructed from one or more
quantum wells within barriers and claddings surrounded by
electrodes. They may also be constructed from aligned "forests" of
quantum rods.
In order to create large externally-applied electric fields,
material layers of the laser device surrounding the
quantum-confined structures (e.g., in a VCSL or VECSEL) may be used
to provide potential differences across the barrier layers and the
quantum well layers. The material and geometric construction
choices should allow the fields to hold the carriers within barrier
layers without significant penetration into cladding layers.
Devices may be electrically or optically pumped. In an electrically
pumped system, n-doped and p-doped cladding material may be used to
inject charge into barrier regions for storage during pumping and
prior to an optical emission phase. Optical pumping schemes may
include directing of optical pumping radiation along a direction
substantially aligned with the electric field provided for
L-switching. For a VECSEL or VCSL configuration, the optical
pumping source may be external to the semiconductor device.
Alternatively, optical pumping radiation may be provided in a
direction substantially orthogonal to the optical emission
direction. For VECSEL or VCSL configurations, optically pumping
radiation may be provided by diodes fabricated on the same
substrate as the switched semiconductor laser.
Details of Exemplary Embodiments
The presented invention provides a significant improvement, by
enabling high peak power emissions to be generated directly from
semiconductors at designed wavelengths selectable over very large
bandwidths. The technology is expected to be an enabler for
wavelength specific detection or material processing applications.
High peak powers improve efficiency in nonlinear processes and
detection ranges. For lidar applications, short pulses provide
range resolution. Wavelengths can be chosen to match specific
transmission, absorption, or scattering peaks of material
constituents.
Technical Approach
Diode lasers have advanced tremendously, with electrical to optical
efficiencies exceeding 70% in some cases. The availability of high
efficiency diodes combined with long lifetime rare-earth crystals
have made Q-switched high-peak-power diode pumped solid state
lasers the standard technology for many applications. For frequency
doubled Q-switched Nd:YAG lasers all the laser power comes through
the diode sources. Currently, diodes themselves are not capable of
achieving high peak powers due to their extremely limited energy
storage capability, with excited lifetimes typically limited to
100's of picoseconds. Consequently, even with highly efficient
diodes used for pumping, the overall efficiencies of these systems
are relatively low (.about.5-10%). Alternatively, a
high-efficiency, high-peak-power diode laser would circumvent the
complexities and inefficiencies associated with optical pumping
solid state crystals and with nonlinear frequency conversion.
We have developed a concept for using an external electric field in
conjunction with geometric designs to actively manage the emission
lifetime.
FIG. 1 shows an exemplary embodiment of a structure and related
band diagram of pumping/emission scheme for producing an L-Switched
light emission as provided for herein. An electric field is
sufficient to separate an electron and hole and suppress
recombination during pumping. To enable lasing, the electric field
is removed.
In a first example of possible embodiments of our concept we
considered core/shell nanorods where an external electric field was
sufficient to "ionize" an electron from an exciton in the core
component of the nanorod. FIG. 1 illustrates the concept for a
CdSe/CdS core shell nanorod. During this ionization stage, the
electron is stored in the shell portion of the rod, and the
recombination rate is vastly reduced. During the "ionized" stage,
the hole may still be confined within the core, or it may also be
"ionized" and mostly residing on the opposite of side of the shell
where the electron is residing. When the electric field is
released, the electron returns to the core, recombines with the
hole, and a photon is emitted. The system has a long lifetime when
the charges are separated and a short lifetime when the charges are
confined within the core layer.
Though the previous example was illustrated for nanorod
nanostructures, the same active control is also applicable to the
excitonic lifetime of layered quantum well nanostructures where
confinement is implemented in only one dimension. These structures
allow for higher exciton densities and are more readily fabricated
with mature processes.
This active switching of the lifetime of a gain medium within a
cavity leads to effects that are very similar to Q-switching in
that energy stored within the gain material is rapidly made
available for lasing, so that the gain exceeds that cavity losses
and a "giant pulse" is emitted. However, the switching is made
through voltage modulation of the gain material itself instead of a
separate electro-optic crystal. We are calling this effect
"L-switching", where the "L" stands for lifetime. Since a
significant Stark-shift in the gain curve of the semiconductor
accompanies the electrical switching, a spectral filter may be
added to the cavity to include modulation of cavity losses as well,
producing combined QL-switching effects.
In the simple form of the device the application of the external
field results in separation of the electron and hole. However, we
have also explored a device concept in which the quantum well is
embedded within a p-n junction. In this case the built-in electric
field at the junction can be used to separate the electrons and
holes. With the p-n junction based device, the external field is
applied to flatten the field across the quantum well and allow for
recombination. The p-n junction may also be used to inject current
into the well, allowing for electrical pumping.
Optical pumping is not an unattractive approach for the described
emitters. Since optical pumping can be performed at wavelengths
near the emitting wavelength, diode pump emitters can be fabricated
within the same chip as the L-switched emitter in an integrated
monolithic component. The overall component would still be
electrically driven.
We have explored the use of II-VI and III-V material systems for
fabrication of blue/green emitting devices. It is worth mentioning
that for many applications, where infrared emission is desired, our
switching process also has merit. For these applications much more
mature device fabrication processes are available based on GaAs
materials.
Lifetime Switching (L-Switching) of a Quantum Well
FIGS. 2 and 3 illustrate a simple scheme implementing L-switching
in a quantum well/barrier nanostructure. FIGS. 2 and 3 show a band
diagram during a pumping/shelving phase and during an emission
phase for an exemplary embodiment of an L-Switched light emission.
Shown are the conduction-band and valence-band potential profiles
along the growth direction of a quantum well/barrier structure.
During the shelving phase an electric field is applied across the
structure, spatially separating the electrons from the holes, which
leads to a suppression of recombination and hence a large excited
population. During the emission phase in FIG. 3, the electric field
is turned off, returning the electron-hole luminescence rate to its
normal value. The shelved excitations are dumped as a strong light
pulse.
During the shelving phase in FIG. 2, an electric field is applied
across the structure. During the shelving phase or part thereof,
the system will be pumped. The system's band structure is designed
in such a way that, under the electric field, the lowest energy
quantum state of the electrons (holes) in the conduction (valence)
band is localized at one end of the barrier, as shown in FIG. 2.
The excited electrons and holes are thus separated from each other
in space, leading to a drastic suppression of both radiative and
non-radiative recombination rates. Under this condition, a large
pumped electron-hole population accumulates over time, and pumping
may occur through optical stimulation or through current injection.
Switching off the electric field starts the emission phase as shown
in FIG. 3. The lowest energy states move back to inside the quantum
well, their wavefunctions overlapping each other spatially. The
radiative recombination rate returns to its normal value, resulting
in the dumping of the shelved excitation energy as a strong light
pulse.
This process can be viewed as a nanoscale capacitive storage of
charge for trigger-able pulsed power photonic emission. We can
compare this approach to pulsed operation of conventional diodes
where energy is stored in external capacitors prior to providing
pulsed injection-currents. Our approach capacitively stores the
charge, as close as quantum-mechanically allowed, prior to
injection into the quantum well. Since the charge migrates over
nanometer-scale distances during emission, the resulting generation
of heat is negligible when compared to diodes where similar
currents are applied over macroscopic distances.
FIGS. 4 and 5 show an exemplary embodiment for producing an
L-switched light emission that is suitable for use with optical
pumping of the structure, including exemplary values for the
geometric length scales, voltages, and electric fields. The example
is designed to avoid electrical break down through fringe fields
around the structure by making the vertical dimension much larger
(shown here as 1.2 mm) compared to the horizontal dimension (shown
here as 120 nm).
L-Switching in a Quantum Well Using a p-n Junction
Though the approach described in FIGS. 2 and 3 can be achieved with
moderate voltages, very large electric fields are required. This is
mainly because the applied field has to be large enough so that the
Coulomb attraction between the separated electrons and holes cannot
dominate over the external field that is trying to separate the
charges.
We investigated a variation of the approach shown in FIGS. 4 and 5,
in which a p-n junction is used to provide the electric field. The
concept of this alternative scheme is sketched in FIGS. 6 and 7.
FIG. 6 shows an exemplary block diagram of a quantum well/p-n
junction structure for producing an L-Switched light emission.
FIGS. 7A and 7B show an exemplary band diagram of a quantum
well/p-n junction structure during a shelving phase and an emission
phase.
The quantum well/barrier structure is embedded in the immediate
neighborhood of the p-n interface in the depletion region on the
p-doped side of a p-n junction of FIG. 6, as designated with the
arrow. The switching is again effected by the external bias
potential, but, in contrast to the simple scheme in FIGS. 2 and 3,
the external field is off during the shelving phase and is turned
on during the emission phase.
The equilibrium charge distribution in the depletion region
provides the needed potential drop across the quantum well/barrier
to separate the electrons and the holes (FIG. 7A). During the
emission phase one turns on the external bias potential, which
offsets the `intrinsic` electric field in the depletion region
(FIG. 7B). The quantum well/barrier being thus returned to a
`flat-band` configuration, the excited electrons and holes move to
overlap each other, resulting in efficient light emission. The
advantage of this scheme is that the two doped segments of the
junction act as a micron-scale capacitor, the potential drop of
which can conveniently be tuned by the low external bias potential.
This enables the design to sidestep the challenge of inserting a
pair of electrodes at the two ends of the nano-scale quantum
well/barrier structure.
Proposed Design of AlN/InGaN L-Switched Light-Emitting
Structure
We have produced a design implementing the L-switching concept
sketched in FIGS. 6 and 7. The materials we have selected are doped
MN for the p-n junction and InGaN for the quantum well and
barriers. We summarize the considerations leading to this choice in
the next subsection. One main design consideration of the
nanostructure is its energy band profile, especially the band
offsets at various interfaces. The band profile depends on two
factors: the material composition in each layer and the charges
accumulated in the junction's depletion region.
FIG. 8 shows a comparison across an exemplary quantum well/p-n
junction structure of band diagrams and strain profile of the
structure. FIG. 8 shows our structure's band diagram before the
formation of the depletion region at the junction, together with
the material parameters used. For the growth orientation assumed
here, no electric fields due to the spontaneous polarization of
III-nitride materials or piezoelectric fields occur. For other
growth directions, these inherent fields can be used to optimize
the structure design with regard to the L-switching requirements.
FIG. 8 also shows the strain profile.
FIG. 9 shows an exemplary band diagram of a quantum well/p-n
junction structure with the effect of charge distribution in the
junction's depletion region. The equilibrium charge distribution,
resulting from charge diffusion between the two doped segments of
the p-n junction and making up the depletion region, produces a
potential gradient across the quantum well/barrier structure, as
shown in FIG. 9 (dashed red line). The potential drop here (about
5.times.10.sup.7 V/m) is generated by a doping density of
10.sup.-18 cm.sup.3 on either side of the junction. Achieving this
doping concentration is presently still challenging for III-nitride
structures, while in conventional (red-emitting) III-V structures
it can be routinely achieved.
FIGS. 10 and 11 show exemplary wavefunctions within a quantum
well/barrier structure during a shelving phase and an emission
phase, respectively. As described in the previous subsection, the
equilibrium potential gradient needs to be sufficiently strong to
separate the pumped electrons and holes during the shelving phase.
FIG. 10 shows that the calculated wavefunctions for these carriers
are indeed well separated by the field generated by the depletion
region charges. The separated charges, in turn, modify the electric
field inside the quantum well/barrier. The energy band profile,
taking into account the effect of a pumped charge density of
10.sup.12 cm.sup.-2, which are approximated as two charged sheets,
is shown as the solid blue line in FIG. 9. It can be seen that this
modification is quite modest. During the emission phase FIG. 11, an
external bias potential of 6.2 V is applied, which compensates for
the potential drop across the junction, returning the bands to a
`flat` configuration, and leading to maximum overlap between the
wavefunctions. The fact that the confinement region comprises two
barriers and a well region is not of crucial importance. We have
also analyzed structures similar to the one shown in FIGS. 10 and
11, but with the well depth reduced to zero, in other words without
a well region in the middle of the confinement region, and found
that those structures can also be used for L-switching.
The luminescence rate is proportional to the square of the absolute
magnitude of the overlap between the electron and hole
wavefunctions. From our wavefunctions, the shelving phase
luminescence rate is calculated to be reduced by a factor of
approximately f.sub.r=10.sup.-10 relative to the emission phase
rate, with
.xi..xi..xi..xi. ##EQU00001##
where sp stands for shelving phase and ep for emission phase, and
.xi..sub.e|.xi..sub.h=.intg.dz.xi..sub.e*(z).xi..sub.h(z).
Materials Selection Considerations
The material composition and dimensions of each layer in the
structure shown in the previous example were selected to have the
required optical and/or electronic properties and to be able to
form interfaces with manageable strain and to provide green light
emission. Green-light emitting L-switch structures can be made from
II-VI and III-nitride materials. GaN based materials may be chosen
because the exploitation of these materials is rapidly maturing and
is already demonstrated in commercial blue-green emitting diodes
and LEDs. Presently, disadvantages of III-nitride materials are
related to the fact that it is difficult to grow high quality
crystals (that exhibit long non-radiative lifetimes), especially in
structures that involve built-in strain, and it is difficult to
dope the structure such that, at room temperature, the free carrier
concentration is close to or larger than 10.sup.17 cm.sup.-3. It is
also still difficult to grow II-VI materials emitting in the green
that would not show degradation problems (i.e. diminishing crystal
quality over time). For L-switching in optically pumped III-nitride
structures, where doped crystal segments are not needed, possible
choices for the cladding material include sapphire or AlN, which
have bandgaps substantially larger than the green emission
frequency and are therefore well-suited as cladding materials. The
quality of the crystal structure of the cladding material is not as
important as that of the barrier material, as the deleterious
non-radiative recombination happens mostly in the barrier material.
In the current embodiment, the dimensions of the quantum
well/barrier were chosen to engineer wavefunctions for green light
transition emissions and sufficient shelving-phase charge
separation.
The concept of L-switching is also valid for III-V compounds
emitting in the red or near infrared. A possible embodiment of an
L-switching structure could be In.sub.xGa.sub.1-xAs as well
material (x to be chosen sufficiently large so that type-I band
alignment at the In.sub.xGa.sub.1-xAs--InP interfaces is ensured,
but not too large so that the well is nearly lattice-matched to the
barrier) and InP as barrier material. In the case of an optically
pumped III-V L-switching structure, a possible cladding material
with sufficiently large bandgap (substantially larger than that of
InP) could be the spinel oxide MgAl.sub.2O.sub.4, which has a
direct gap of approximately 5.36 eV. For a p-n junction embodiment
of a III-V L-switching structure, p- and n-doped cladding materials
might be chosen from the II-VI material system, for example CdS or
CdSe.sub.yS.sub.1-y, latticed matched to InP.
The concept of L-switching is also valid for organic light-emitting
diodes (involving organic p-n junctions) and light-emitting
electrochemical cells. Charge carrier mobilities are generally
smaller in organic materials compared to inorganic crystals. This
may affect in particular the carrier capture process, described
further herein, possibly leading to slower carrier capture and
carrier recombination in the emission phase of the L-switching
process, hence increasing the lower limit of the pulse duration in
organic vs inorganic L-switching structures.
Non-Radiative Losses of Pumped Electron-Hole Populations
To attain the goal of long-lifetime energy storage, our invention
reduces the luminescence rate by separating the pumped charges in
the quantum well. However, this goal may still be compromised by
losses of excitations via non-radiative recombination processes.
Major non-radiative losses are through (i) the Shockley-Read-Hall
(SRH) process, where bulk and surface defects spatially trap the
electrons and the holes which then recombine, dissipating the
excitation energy as heat, and (ii) Auger recombination where an
electron-hole pair recombines, transferring their excitation energy
to another charge instead of releasing it as a photon. These two
processes are illustrated in FIGS. 12 and 13. Our analysis shows
that during the shelving phase, these non-radiative processes are
also substantially suppressed by charge separation so that even in
their presence, microsecond-scale lifetimes should be
achievable.
FIG. 12 shows an exemplary band diagram of a quantum well/barrier
structure and non-radiative Shockley-Read-Hall (SRH) recombination
processes therein. In the SRH process, the recombination is
facilitated by trapping of the charges by impurities. FIG. 13 shows
an exemplary energy diagram showing Auger recombination processes
therein. In the illustrated Auger process, an electron (in level
C1) and a hole (in level V1) recombine, transferring their energy
via Coulomb to another C1 electron, which is excited to level
C2.
FIGS. 14 and 15 show an exemplary recombination factor for
non-radiative Shockley-Read-Hall (SRH) recombination processes
within a quantum well/barrier structure during the shelving phase
and the emission phase, respectively. The factor F.sub.SRH(z) is
shown therein as a function of physical position during,
respectively, the shelving phase and the emission phase. During the
shelving phase shown in FIG. 14, the factor's peak value is reduced
by at least a factor of 10.sup.-5 as compared to the emission phase
shown in FIG. 15.
SRH Recombination
The trapping impurities responsible for the SRH process may reside
in the `bulk` of a layer or in an interface. The general theory of
the SRH process gives the bulk recombination rate as
.apprxeq..times..times..intg..function..times..times..function..function.-
.times..function..function..function. ##EQU00002##
n.sub.e(z) and n.sub.h (z) are the electron and hole densities
respectively, L.sub.z is the total width of the quantum
well/barrier structure, and w.sub.nr.sup.(0,bulk) is a constant
characteristic of the material and the impurity density. The
interface SRH rate is similarly given by
w.sub.nr.apprxeq.w.sub.nr.sup.(0,interface)F.sub.SRH(z.sub.i),
where z.sub.i is the location of the interface, and
w.sub.nr.sup.(0,interface) is the interface characteristic scale
constant. Since both bulk and interface rates depend on the factor
F.sub.SRH(z), which depends on the overlap of the electron and hole
densities, one expects the SRH recombination rates to also be
suppressed by the charge separation during the shelving phase. The
factor F.sub.SRH(z), calculated from our wavefunctions, is shown in
FIGS. 14 and 15 to be reduced by several orders of magnitude during
the shelved configuration. Using these result, we have found that,
for the bulk SRH rate, we have nonradiative recombination rate
suppression ratios:
.times. ##EQU00003## and for the interface rate.
.apprxeq..times. ##EQU00004## Auger Recombination
For the analysis of the Auger recombination rate, we may write the
quantum well electron wavefunction in the form
.phi..function..times..times..xi..function..times..function.
##EQU00005##
where u.sub.vk (r, z) is the lattice-periodic unit-cell part of the
Bloch wavefunction with band label v and crystal momentum k, r
represents the in-plane spatial coordinates, .xi..sub.vl (z) is the
quantum well envelope function of subband l, and A is a
normalization constant. The Auger recombination rate is
proportional to the (absolute) square of the Coulomb scattering
matrix element between two electrons, which is given by
V.sub.1,2,3,4.sup.c.apprxeq.V.sub.k.sub.1.sub.-k.sub.3.sup.cF.sub.1,2,-
3,4(|k.sub.1-k.sub.3|).delta..sub.k.sub.1.sub.-k.sub.3.sub.,k.sub.4.sub.-k-
.sub.2u.sub.v.sub.1.sub.k.sub.1u.sub.v.sub.3.sub.k.sub.3.sub.cellu.sub.v.s-
ub.2.sub.k.sub.2u.sub.v.sub.4.sub.k.sub.4.sub.cell with
F.sub.1,2,3,4(q)=.intg.dzdz'.xi..sub.v.sub.1.sub.l.sub.1*(z).xi..sub.v.su-
b.2.sub.l.sub.2*(z')e.sup.-q|z-z'|.xi..sub.v.sub.3.sub.l.sub.3(z).xi..sub.-
v.sub.4.sub.l.sub.4(z')
For the Auger process illustrated in FIG. 13, we can take, as an
example, v.sub.1=V1, v.sub.2=C2, v.sub.3=C1, v.sub.4=C1. This
matrix element also depends strongly on the overlap of the electron
and hole wavefunctions. For q=0 it contains a factor identical to
the overlap matrix element that controls the radiative decay. Hence
the Auger process is suppressed by charge separation similarly
efficiently as the optical transition process during the shelving
phase.
Carrier Capture
The L-switching processes involves complex dynamics of the charge
distribution and, during the E-field switch-off process, local
charge recombination of the initially spatially separated charges.
The charge recombination is accompanied by a carrier capture into
the well (FIGS. 16A and 16B). The time scale of this capture
process depends on various variables, including the system's
temperature, excitation density, and the details of the band
profile. The time scale is expected to be in the picosecond regime.
FIG. 16A shows a schematic of an electron wavefunction during
shelving and FIG. 16B shows a schematic of the electron capture
after switching. Therein, the shelving phase of FIG. 16A may serve
as an approximate initial condition for scattering and capture in
the unbiased structure of FIG. 16B.
Leakage Due to Tunneling
Another important aspect of the structure design is the importance
of the large confinement barriers provided by the cladding
material. The cladding barrier, .DELTA..sub.cladding, may be chosen
high enough (for both electrons and holes) such that carrier
leakage due to tunneling (FIGS. 17A and 17B) will not degrade the
device performance. FIG. 17A shows a band diagram schematic of
carrier leakage due to tunneling in an exemplary structure. FIG.
17B shows estimated calculations of tunneling time in such an
exemplary structure.
Initial simplified estimates for the tunneling time based on a
kinetic model involving a round-trip time and a transmission
coefficient T yield the following expression for the tunneling time
(subscript tn) .tau..sub.tn.apprxeq..tau..sub.rt/T, where
.tau..sub.rt is the round-trip time in the well (here
barrier/well/barrier) section, and
.apprxeq..times..times. .times..times..DELTA. ##EQU00006## and with
l being the effective tunneling distance. Preliminary estimates
shown in FIG. 17B show that the design shown in FIGS. 4 and 5 will
not suffer from leakage problems. However, the design in FIGS. 4
and 5 assumes an AlN or sapphire cladding, and a sapphire cladding
would provide a particularly large confinement potential, but may
lead to lower crystal quality of the III-nitride segments than an
AlN cladding.
For a pump duration of 1 microsecond, a barrier of approximately
0.7 eV is needed. The design shown in FIGS. 4 and 5 substantially
surpasses this requirement for both electrons and holes and can
therefore be expected to perform without leakage problems. For
smaller cladding barriers, provided for example by AlN cladding, a
more realistic theory for the tunneling is needed in order to
obtain precise predictions of the tunneling process and to prevent
leakage problems during the pumping/shelving phase of the
L-switching process.
A rate-equation model may be used to simulate the operation of an
L-switched laser. The following describes the predicted performance
of an L-switched laser composed of the pumped light emitting unit
described in the previous section placed inside a resonance cavity.
The model is described in detail below.
This model is an adaptation of an established model, used for
lasers with bulk semiconductor emitters, to the two-dimensional,
quantum well setting. The model equations govern the evolution of
the (planar) densities of pumped electrons, N.sub.e, and holes,
N.sub.h, which are assumed to be equal (N.sub.e=N.sub.h.ident.N),
the electron density in the active spectral range (e.g., lasing
window), n.sub.e, (see FIG. 18) the hole density in the lasing
window, n.sub.h, and the (planar) cavity photon density, S.
FIG. 18 shows an exemplary embodiment of band diagram showing
carrier energies as functions of momentum in the plane of the
quantum well. Momentum conservation restricts optical transitions
to between states with almost equal momenta. The spectral ranges in
the two bands in resonance with the laser constitute a "lasing
window."
We show here the results of a simulation of the L-switch operation
in the quantum well laser. For the lasing mode, we chose the
wavelength .lamda.=532 nm and the linewidth .delta..lamda.=6 nm.
The other parameter values are set as follows: m.sub.e=0.22
m.sub.0, m.sub.h=0.40 m.sub.0, where m.sub.0 is the free electron
mass, .tau..sub.s=1.0.times.10.sup.-10 s,
.tau..sub.1e=.tau..sub.1h=1.0.times.10.sup.-12 s,
.tau..sub.p=8.3.times.10.sup.-10 s, r.sup.2=(5.times.10.sup.-10
m).sup.2, |.PHI.(z.sub.QW)|.sup.2=10.sup.7 m.sup.-1,
.alpha..sub.f=0, .beta.=0.01, .beta.'.sub.2=0, .GAMMA.=1,
.epsilon..sub.cav=6.25, T=298 K, .alpha..sub.cav=0. The
non-radiative lifetime .tau..sub.s and the optical transition
dipole squared r.sup.2, to which all the emission rates are
proportional, are set to their `normal` values, i.e. the values
under the flat-band condition in the emission phase (FIG. 11),
where the electron and hole wavefunctions overlap maximally. During
the shelving phase in the simulation, to account for the effect of
wavefunction (charge) separation, .tau..sub.s is lengthened by a
factor of 1/f.sub.s=10.sup.6 while r.sup.2 is reduced by a factor
of f.sub.r=10.sup.-10:
r.sup.2(t)=[1-s(t)]r.sup.2+s(t)f.sub.rr.sup.2
.tau..sub.s(t)=[1-s(t)].tau..sub.s+s(t)f.sub.s.tau..sub.s
with the smooth switching function
.function. ##EQU00007##
For each laser pulse, the length of the shelving phase is set at
10.sup.-6 s, during which the pump current is kept on. At the end
of this period, t.sub.dump, the parameters .tau..sub.s and r.sup.2
are smoothly changed to their emission phase values of a switching
duration of t.sub.s=10.sup.-8 s. The calculated results for the
various densities for a pump current density of 10.sup.4 A/m.sup.2
and a quantum well area of 1 cm.sup.2 are shown in FIGS. 19, 20A
and 20B. FIG. 19 shows output power of a light pulse from an
exemplary L-switched laser as described. The total energy density
in the pulse is approximately 0.01 J/m.sup.2. FIG. 20A shows
shelved electron-hole density, N, as a function of time. FIG. 20B
shows population inversion, 1-f.sub.e-f.sub.h, where f.sub.e and
f.sub.h are the averaged electron and hole occupation probability,
respectively, in the lasing window.
It can be seen in FIG. 20A that up to 6.times.10.sup.12
electron-hole pairs per cm.sup.2 are stored and in FIG. 19 that the
energy is released as a 5-ns laser pulse with an energy density of
0.01 J/m.sup.2. If we assume the electron-hole pairs are injected
at the lasing frequency, the internal efficiency of this simulation
run is given by:
.times..times..times..times..times..times. ##EQU00008##
A more detailed discussion of efficiencies appears in the next
subsection.
FIGS. 21A and 21B show additional detail of cavity photon density
and electron/hole pair density and population inversion at the
laser frequency around the switching time. The results are plotted
in more details over a 100-ns interval around the switching time
t.sub.dump. It can be seen that lasing stops when the shelved
electron-hole density drops to a threshold level where the
corresponding inversion in the lasing window is near zero. The
pulse duration is determined by the model parameters given above
that govern the stimulated emission process, as well as the
switching function s(t). The pulse duration of about 5 ns shown in
FIGS. 19, 20A and 20B, where a switch duration t.sub.s of 10 ns has
been used, does not represent the theoretical minimum for
achievable pulse durations.
We define the internal efficiency of our model L-switched quantum
well laser by
.eta..times..times..times..times..times. ##EQU00009##
where
.times..times..tau..times..intg..times..times..function.
##EQU00010## is the number of photons per unit area in the output
laser pulse, and I.sub.abs is the pump current absorbed by the
quantum well/electronic charge/unit area.
For the sake of analysis, it is useful to write this efficiency as
a product of two component efficiencies:
.eta..sub.q=.eta..sub.pump.eta..sub.output.
The pump efficiency is defined as
.eta..times..times..times..times..times. ##EQU00011## where
N.sub.max is the maximum shelved electron-hole density immediately
before the switching, and the output efficiency is defined as
.eta..times..times..times..times. ##EQU00012##
The output efficiency measures how much of the stored energy in the
quantum well during the shelving phase is converted into output
laser energy. We recall that (FIG. 21) lasing stops when N drops to
a level, which we now denote by N.sub.th, when the inversion in the
lasing window is near zero. Therefore, the maximum energy density
that can be extracted, even when all dissipation losses can be
neglected, is N.sub.max-N.sub.th, which gives an upper bound to the
output efficiency:
.eta..ltoreq..times..times..times..times. ##EQU00013## We plot the
inversion at the lasing frequency against the shelved carrier
density for the parameter values chosen in the simulation in FIG.
22, where we can read out a value for the threshold level
N.sub.th.apprxeq.2.6.times.10.sup.12 cm.sup.-2, which, with
N.sub.max.apprxeq.6.times.10.sup.12 cm.sup.-2, gives
.eta..sub.output.ltoreq.0.56. FIG. 22 shows a graph of exemplary
population inversion at the laser frequency versus the shelved
electron/hole pair density.
FIG. 23 shows a graph of the dependence of the pump efficiency on
the switching time t.sub.dump. Therein, .tau. is the decay time of
the shelved carrier density.
For the pump efficiency, a simple analytic estimate can be obtained
as follows. We approximate the equation for the electron-hole
density during the shelving phase as
.apprxeq..times..times..tau. ##EQU00014## where .tau. is the decay
time for N. This gives N(t)=.tau.I.sub.abs (1-e.sup.-t/.tau.).
Since N.sub.max=N(t.sub.dump), an estimate of the pump efficiency
can be obtained:
.eta..times..times..times..times..times..apprxeq..tau..times..tau.
##EQU00015## This estimated value for this efficiency is plotted in
FIG. 23. We note that this efficiency is independent of I.sub.abs
and N.sub.max.
Combining the bounds on .eta..sub.pump and .eta..sub.output, we
obtain an estimated bound on the internal efficiency
.eta..eta..times..eta..apprxeq..times..times..times..times..times..tau..t-
imes..tau..times..tau..times..tau. ##EQU00016##
It can be seen from FIG. 23 that the .eta..sub.pump is higher for
smaller t.sub.dump. Since the desirable level of N.sub.max is set
by other considerations, e.g. .eta..sub.output and the electric
field generated by the separated charges, one can adjust the
absorbed input current to reduce t.sub.dump while keeping N.sub.max
at a favored level.
FIG. 24A shows a graph of switching time versus absorbed current
density whereas FIG. 24B shows analytic estimates of efficiencies
for the noted efficiencies for an L-switched emission versus
absorbed current density (solid lines), as compared with simulation
results (squares).
In FIG. 24A, we show this dependence of t.sub.dump on I.sub.abs for
a desired target density of N.sub.max=6.times.10.sup.12 cm.sup.-2
and .tau.=10.sup.-4 s that follows from
.function..times..times..tau..times..times..times..times..function..tau.
##EQU00017##
FIG. 24B shows, for the same level of N.sub.max, the internal
efficiency as estimated by the above equation. It agrees very well
with the computed value from the results of our simulations.
We conclude with a comparison of the predicted performance
characteristics of our simulated L-switched laser with exemplary
target specifications. The target delivered power is taken to be 50
W at 400 pulses/s, or an energy of 125 mJ/pulse. With
N.sub.max=6.times.10.sup.12 cm.sup.-2 and .eta..sub.output=0.56,
the maximum energy per pulse from our laser is 13 mJ/m.sup.2.
Power Scaling
We present three scaling mechanisms that may be used to increase
the overall output power: multiple wells per emitter, multiple
emitters per laser cavity, and incoherent beam combining of
multiple cavities.
FIGS. 25A and 25B show top and side views of a concept for single
emitting element composed for multiple layers of switched quantum
wells. In this arrangement, interdigitized conductive layers will
be used to maintain separation between alternating voltage
polarities and avoid electrical breakdown. The electrodes consist
of thin films of Indium Tin Oxide (ITO), or other transparent
conductive materials, between each layer and the polarities of the
p-n junctions are alternating between electrodes. The component may
be cooled through thermally conductive windows (e.g. Sapphire or
SiC) on both sides of the element.
An estimate on the thickness of each layer is 5 .mu.m or more, and
one may project that 200 layers of switched quantum layers may be
constructed within a 1 mm thick integrated component. This
thickness is sufficient to limit the risk of electrical breakdown
between layers along the sides of the device. Though one side of
the device could include a grown distributed Bragg reflector, as
typically used in VECSELs, transparent top and bottom windows may
allow for more compact arrangements of multiple devices in a single
laser cavity.
However, placement of the device with respect to the standing modes
of the cavity may become more of a challenge in that case. To
assure reasonable overlap between quantum wells and antinodes, a
diversity of quantum well spacings may be included within the
design. Alternatively or additionally, the laser may be configured
in a ring-cavity.
One may also use multiple emitting elements within a single cavity
as shown in FIGS. 26A and 26B. In this diagram, the cavity
traverses 9 separate gain elements. These may be grown on a single
substrate. Provided that these components can be made to support
high transmission surfaces, a relatively compact serpentine cavity
may be made using turn mirrors, as depicted in the side view. A
slightly less compact version of the concept could be used if the
bottom surfaces of the gain elements are constructed with DBR
reflectors.
Using a concave-convex mirror arrangement results in modest
variation in beam size within the cavity and enables high overlap
efficiency with the individual gain elements. In the top view, the
positive and negative electrodes are shown that engage the
interdigitized conductive layers within each gain region. Since
fast triggering times will be needed, the capacitance between the
positive and negative polarity electrodes should be minimized. One
way to do this is to limit the ITO conduct layers to the regions
where gain will be created and mask the ITO layers so that
conduction is not allowed between emitters.
Beam Combining with Arrays and Tiers
Multiple L-switched cavities may be arranged in arrays and
synchronously triggered. FIG. 27A illustrates an array of cavities
that may be integrated within a single integrated layer as a
synchronized linear array.
The concept here is that multiple iterations of the architecture
shown in FIGS. 26A and 26B may be assembled as a single integrated
component. In this arrangement, the electrodes for each successive
cavity may be reversed in polarity so that the electrodes between
each cavity may be shared if necessary. For applications where
illumination at range is required, it will be important to design
and calibrate the output lens alignment so to minimize or control
the divergence in beam directions. However, for many applications,
where a large aperture is not required, incoherent beam combining
from multiple elements may be adequate.
Given the architecture shown in FIG. 26A for each cavity, a tightly
packed array of emission beam lines can be achieved by stacking the
linear arrays in displaced tiers as shown in FIG. 27B.
This arrangement results in a compact N.sub.layer.times.N.sub.tier
array of aligned and synchronized pulsed laser that is suitable for
line illumination in applications that do not have stressing
divergence requirements. It is also worth noting that such a laser
array may be used as a pump source for pulsed solid-state gain
materials having short upper state lifetimes and requiring short
pump pulse durations.
One possible embodiment of the invention includes the following
specifications for a 3 mm radius beam waist within a cavity and
proving a total quantum well area A.sub.QW=2.83.times.10.sup.-5
m.sup.2:
Use N.sub.wells=200 Quantum Wells per Stack
Use N.sub.stacks=24 Stacks per Oscillator
Use N.sub.Osc=15 Oscillators per Tier
Use N.sub.Tiers=5 Tiers in an Array
This embodiment uses quantum well stacks, oscillators, and tiers
fabricated in a tightly integrated and reproducible manner. If we
assume that each of stacked quantum well devices requires 1.0
cm.sup.3 (including surrounding roof mirrors and packaging space),
the overall optical system might end up being about
10''.times.6''.times.2''.
Additional Considerations
Though most of the above description has been tied to applications
for high peak power green laser emission with an InGaN material
system, this invention includes the described processes and methods
applied to other material systems and architectures. Materials may
be based on organic semiconductors or inorganic semiconductors.
Gain elements may include one or more quantum confined volumes in
physical contact with barrier materials such that an external
electric field moves electrons or holes substantially into the
barrier materials to reduce the probability of recombination. A
cladding material may be used between the barrier material and
electrode surfaces. The cladding materials may be arranged to form
a p-n junction to enable current injection into the barrier or
quantum well region, or to provide electric field control for
switching the recombination lifetime within one or more quantum
wells. The quantum confined volumes may be shaped for quantum
confinement in one, two, or three dimensions (quantum wells, wires,
or dots).
There are many different configurations of the described
embodiment, here laid out in specificity. FIG. 28 shows an
exemplary structure for zero-dimensional and one-dimensional
quantum confined volumes. One or more quantum wells may be inserted
between opposite polarity electrodes. Multiple alternating-polarity
electrodes may be used to integrate multiple emission regions
within a single integrated gain element. FIG. 29 shows an exemplary
gain stack with multiple quantum wells per electrode pair.
The gain elements may be transparent or include one or two
integrated reflective surfaces. FIG. 30 shows an exemplary
configuration for using multiple transmissive gain stacks in
succession. FIG. 31 shows an exemplary configuration for using
multiple reflective gain stacks in succession.
The described gain elements may be optically or electrically
pumped. For optical pumping, the optical pump source may be
external to the device or may be included as one or more integrated
diode optical sources.
FIG. 32 shows an exemplary configuration for using side-pumping
diodes with an integrated switched gain element. In FIG. 32, the HR
surfaces are highly reflective surfaces for the pump diodes which
are directed through output coupler interfaces (not shown) towards
the quantum well structures within the switched laser region of the
device.
FIG. 33 shows an exemplary configuration for using external optical
pumping with a gain element. One or more of the described gain
elements may be inserted within an optical cavity. A spectral
filter within the cavity may additional inhibit stimulated emission
at the stark-shifted emission wavelength(s) of the gain elements
during the pumping stage.
A Laser Model of Quantum Well Lasers
A semiconductor quantum well laser model has been developed to
begin exploring electrical modification of lifetimes through
electrostatic charge separation in laser systems. This model is
adapted from theory presented in open literature. The model is
briefly discussed in this section without detailed derivations. We
have implemented the model in a Matlab code, initially with some of
the parameters turned off (two-photon absorption and free carrier
absorption). The confinement factor is currently set to unity for
consistency with a VECSEL type architecture.
This model tracks the dynamics of a 2-D carrier density (N). The
total electron and hole carrier densities are equal to preserve
charge neutrality. However, the portion of the electron and hole
distributions that provide energy transitions within the laser
cavity mode linewidth are segregated as active carrier conduction
and valence band densities, n.sub.e and n.sub.h. A cavity photon
density, S, is also included within the set of rate equations.
The model currently includes the following input parameters:
TABLE-US-00001 I = injection current (Amps) A = QW Area (m.sup.2)
.GAMMA. = confinement factor (unitless) m.sub.e = electron
effective mass (kg) m.sub.h = hole effective mass (kg)
.omega..sub.0 = optical frequency of laser (Hz) .delta..omega. =
laser line width (Hz) T = Temperature (K) .alpha..sub.f = free
carrier absorption (m.sup.2/s) .beta. = proportion of emissions in
lasing mode .GAMMA..sub.2 = Two photon absorption parameter
.beta..sub.2' = two photon absorption coefficient .tau..sub.s =
non-radiative decay time (s) .tau..sub.1h = local k-distribution
relaxation time for holes (s) .tau..sub.1e = local k-distribution
relaxation time for electrons (s) E.sub.g = Semiconductor Bandgap
(J) .PHI.(z.sub.QW) = photon mode wavefunction at QW (m.sup.-1/2)
1/.tau..sub.p = radiative loss (1/s) .alpha..sub.cav = other cavity
losses (1/s) .sub.cav = Dielectric constant of cavity material
(unitless)
Additionally, the dipole integral r is pre-calculated from the
quantum well wavefunctions. For simulations where electrons and
holes are electrostatically separated prior to optical switching,
this parameter is provided as a function of a dynamic externally
applied electric field.
The rate equations that are solved are listed below, and some
descriptions of the most significant parameters will follow:
.tau..function..times..alpha..times..GAMMA..GAMMA..times..beta.'.times..t-
imes..times..GAMMA..times..times..function..alpha..times..tau..alpha..time-
s..beta..times..times.'.times..times..beta.'.times..times..times..tau..tim-
es..times..function..times..alpha..times.'.times..times..times..times..tim-
es..times..tau..times..times..times..times..alpha..times.'.times..times..t-
imes..times..times..times. ##EQU00018##
Several intermediate parameters are calculated from the input
parameters:
The thermal parameter
.beta..times. ##EQU00019## is convenient in the carrier
distribution calculations
The exciton reduced mass:
.gamma. ##EQU00020##
The active region material index of refraction: n.sub.b= {square
root over ( .sub.cav)}
The in-band kinetic energies:
.times..times..gamma..times. .omega..times..times..times..times.
.times..times..gamma..times. .omega. ##EQU00021##
The in-band density of states:
.delta..omega..times..times..gamma..times. ##EQU00022##
The laser stimulated emission rate coefficient is proportional to
the transition dipole integral through the following
expression:
.times..omega..delta..omega..times. .times.
.times..times..PHI..function. ##EQU00023##
Likewise, the spontaneous emission rate is also dependent on the
transition dipole integral:
.alpha..function..times. .times..times..times..alpha.
##EQU00024##
In general, the spontaneous emission coefficient for the complete
carrier density (N) is calculated with an integral over the
electron and hole populations:
.pi..times..times..times..intg..infin..times..times..times..function.
.mu..times..function. .mu. ##EQU00025##
where the Fermi-function is given by
.function. .mu..function. .mu..times. ##EQU00026##
In our simulation, we are currently using an approximation to this
integral.
.apprxeq..times..function..times. .times. .times..times..times.
.times..times..pi..times..times. ##EQU00027##
The partial electron and hole populations (n.sub.e and n.sub.h)
experience loss from spontaneous emission that is scaled for the
proportion of the populations that interact with the laser cavity
mode:
'.times. ##EQU00028##
There are several parameters within the rate equations that have a
dependence on the complete carrier density N and act as dynamic
coupling variables. Specifically the Fermi-densities
n.sub.e=N.sub.0 f( .sub.0e,.mu..sub.e,T) and n.sub.h=N.sub.0f(
.sub.0h,.mu..sub.h,T) are dependent on the electrochemical
potentials .mu..sub.e=k.sub.BT ln[exp(.beta. .sub.Fe)-1] and
.mu..sub.h=k.sub.BT ln[exp(.beta. .sub.Fh)-1], which are dependent
on the carrier density through the energy levels
.times..times..times..pi..times..times..times..times..times..times.
.times..times..times..pi..times..times. ##EQU00029##
This patent description and drawings are illustrative and are not
to be construed as limiting. It is clear that many modifications
and variations of this embodiment can be made by one skilled in the
art without departing from the spirit of the novel art of this
disclosure. While specific parameters, including doping, device
configurations, parameters of components, and thresholds may have
been disclosed, other reference points can also be used. These
modifications and variations do not depart from the broader spirit
and scope of the present disclosure, and the examples cited here
are illustrative rather than limiting.
* * * * *